CO2 foam huff and puff reservoir particle monitoring method based on intelligent tracing and dual-mode sensing

Abstract

The invention provides an intelligent tracing and dual-mode sensing CO2 foam huff and puff reservoir particle monitoring method which comprises the following steps: mixing intelligent tracing particles and a CO2 foam oil-displacing agent according to a preset proportion, and injecting the mixture into a reservoir through an injection well; the intelligent tracing particles are monitored through a dual-mode sensing monitoring device arranged at a reservoir monitoring position, and an electromagnetic signal and a fluorescence signal from reservoir fluid are obtained; pre-processing to obtain a pre-processed fluorescence signal and a pre-processed electromagnetic signal; carrying out fusion analysis on the preprocessed fluorescence signal and the preprocessed electromagnetic signal by adopting a data fusion algorithm to obtain current monitoring data of the reservoir particles; according to the method, accurate and real-time monitoring of the current concentration and the current migration state of reservoir particles in a high-temperature, high-pressure and strong-interference complex environment is realized by adopting a layering cooperation scheme of near-well fluorescence microscopic monitoring and far-field magnetic signal macroscopic positioning, data support is provided for reservoir physical property time-varying rule prediction, and the oil and gas exploitation efficiency and stability are improved.

Description

Technical Field

[0001] This invention relates to the field of oil and gas field development technology, and in particular to a method for monitoring particulate matter in CO2 foam huff and puff reservoirs using intelligent tracing and dual-mode sensing. Background Technology

[0002] Medium-deep extra-heavy oil reservoirs are characterized by their deep burial depth and extremely high viscosity. Conventional thermal recovery faces challenges such as severe heat loss and limited vapor propagation range. Furthermore, the disordered migration of particles within reservoir pores further exacerbates reservoir heterogeneity, inducing vapor channeling and leading to the enrichment of remaining oil, significantly increasing the difficulty of its utilization. CO2 foam huff and puff technology, leveraging the viscosity-reducing properties of CO2 and the advantages of foam profile control, can both suppress channeling and improve thermal recovery efficiency to a certain extent, while also aligning with carbon emission reduction goals, thus constructing a synergistic paradigm of "oil enhancement" and "carbon sequestration." However, the profile control effect and channeling prevention accuracy of this technology entirely depend on the real-time and precise understanding of particle concentration and migration status. Only by clearly understanding the particle migration trajectory, retention location, and channel evolution patterns can injection and production parameters be optimized and foam injection intensity adjusted to block vapor channeling paths at the source.

[0003] However, existing monitoring technologies have obvious shortcomings: acoustic sensors can only monitor concentration but cannot capture the migration state, making it difficult to predict the formation trend of cross-flow channels; machine vision sensors are affected by heavy oil adhesion and cannot simultaneously and accurately acquire core indicators; laboratory testing data is lagging and cannot meet the needs of real-time prevention and control. None of these technologies can support the accurate prediction and effective control of cross-flow.

[0004] Therefore, how to capture the dynamics of reservoir particles through precise and efficient monitoring technology, provide data support for the optimization of CO2 foam huff and puff technology, and thus improve mining efficiency and stability, has become a key technical problem that needs to be solved by those skilled in the art. Summary of the Invention

[0005] This invention provides a method for monitoring reservoir particulates using intelligent tracing and dual-mode sensing in CO2 foam huff and puff, which enables accurate and efficient monitoring of reservoir particulate dynamics, thereby improving extraction efficiency and stability.

[0006] On one hand, the present invention provides a method for monitoring particulate matter in CO2 foam huff and puff reservoirs using intelligent tracing and dual-mode sensing, comprising: The intelligent tracer particles are mixed with CO2 foam oil displacement agent in a preset ratio and then injected into the reservoir through an injection well; wherein, the intelligent tracer particles include tracers that can emit fluorescent and magnetic signals; The intelligent tracer particles are monitored by a dual-mode sensing monitoring device deployed at the reservoir monitoring location to obtain electromagnetic and fluorescence signals from the reservoir fluid. The fluorescence signal and the electromagnetic signal are preprocessed to obtain a preprocessed fluorescence signal and a preprocessed electromagnetic signal. The preprocessed fluorescence signal and the preprocessed electromagnetic signal are fused and analyzed using a data fusion algorithm to obtain the current monitoring data of reservoir particles; the current monitoring data of reservoir particles includes the current concentration of reservoir particles and the current migration state of reservoir particles.

[0007] The present invention provides a method for monitoring reservoir particulates using intelligent tracer and dual-mode sensing in CO2 foam injection. This method involves mixing intelligent tracer particles, which possess both fluorescence and magnetic signal emission capabilities, with a CO2 foam oil displacement agent in a predetermined ratio, and then injecting the mixture into the reservoir through an injection well. Dual-mode sensing devices deployed at reservoir monitoring locations collect electromagnetic and fluorescence signals from the reservoir fluid in real time. After preprocessing the signals, including filtering, amplification, and baseline correction, a data fusion algorithm is used to perform collaborative analysis of the preprocessed dual signals. This method achieves accurate and real-time monitoring of the current concentration and migration state of reservoir particulates under complex environments with high temperature, high pressure, and strong interference. This provides data support for predicting the time-varying patterns of reservoir properties, thereby improving extraction efficiency and stability. Attached Figure Description

[0008] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0009] Figure 1 This is a flowchart illustrating the intelligent tracing and dual-mode sensing method for monitoring particulate matter in CO2 foam huff and puff reservoirs provided in this embodiment of the invention. Figure 2 This is a schematic diagram of the structure of the intelligent tracing and dual-mode sensing CO2 foam throughput reservoir particulate monitoring system provided in an embodiment of the present invention. Detailed Implementation

[0010] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0011] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this invention, terms such as "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0012] Figure 1 This is a schematic flowchart of the CO2 foam huff and puff reservoir particulate monitoring method provided in this embodiment of the invention.

[0013] like Figure 1 As shown in the embodiments of the present invention, the intelligent tracing and dual-mode sensing method for monitoring particulate matter in CO2 foam huff and puff reservoirs mainly includes the following steps: 101. After mixing the intelligent tracer microparticles with the CO2 foam oil displacement agent in a preset ratio, the mixture is injected into the reservoir through an injection well; In one specific implementation process, the intelligent tracer microparticles employ a fluorescent-magnetic composite nano-tracer. This tracer has a core-shell structure; the core emits a stable magnetic signal, while the outer shell emits a specific fluorescent signal. Both signals exhibit good high-temperature and high-pressure tolerance, making it suitable for the complex operating conditions of reservoirs ranging from 50-300℃ and 10-30MPa. Before initiating CO2 foam huff and puff operations in heavy oil, a preset mixing ratio of the intelligent tracer microparticles and the CO2 foam displacement agent can be determined based on reservoir geological parameters (such as porosity and permeability) and production design requirements. This ensures that the tracer microparticles are uniformly dispersed in the displacement agent, and that the signal strength meets monitoring requirements. Subsequently, the mixed intelligent tracer microparticles and CO2 foam displacement agent are injected into the reservoir through injection wells. During injection, the injection pressure and injection rate are controlled to prevent premature particle aggregation due to excessively rapid fluid erosion.

[0014] 102. The intelligent tracer particles are monitored by a dual-mode sensing monitoring device deployed at the reservoir monitoring location to obtain electromagnetic and fluorescence signals from the reservoir fluid. In a specific implementation, the dual-mode sensing monitoring device consists of a fluorescence sensing module and an electromagnetic sensing module, arranged in a concentric circle + radial array at preset monitoring locations in the reservoir. This arrangement provides omnidirectional coverage of the reservoir, ensuring the capture of the entire trajectory of particle migration. The fluorescence sensing module contains multiple fluorescence sensor nodes to receive fluorescence signals emitted by the intelligent tracer particles; the electromagnetic sensing module contains multiple electromagnetic sensor nodes to capture magnetic signals generated by the particles. When the intelligent tracer particles diffuse and migrate within the reservoir along with the CO2 foam displacement agent, the dual-mode sensing monitoring device initiates monitoring at a preset data acquisition frequency, simultaneously acquiring electromagnetic and fluorescence signals from the reservoir fluid.

[0015] 103. The fluorescence signal and the electromagnetic signal are preprocessed to obtain a preprocessed fluorescence signal and a preprocessed electromagnetic signal; In a specific implementation, the preprocessing process can include three steps: filtering, signal amplification, and baseline correction. Filtering removes irrelevant interference signals such as CO2 foam vibration and equipment operating noise, while retaining effective signals related to intelligent tracer particles. Signal amplification addresses the issue of weak particle signals deep in the reservoir by using amplification circuits to enhance the strength of effective signals and ensure the reliability of subsequent analysis. Baseline correction eliminates signal drift caused by factors such as ambient temperature and pressure fluctuations. It uses the signal baseline at the initial monitoring time as a reference to calibrate subsequently acquired signals and ensure signal stability.

[0016] 104. The preprocessed fluorescence signal and the preprocessed electromagnetic signal are fused and analyzed using a data fusion algorithm to obtain the current monitoring data of reservoir particles; the current monitoring data of reservoir particles includes the current concentration of reservoir particles and the current migration state of reservoir particles.

[0017] In a specific implementation, the core of the data fusion algorithm is to combine the advantages of two signals to overcome the limitations of a single signal: fluorescence signals have high sensitivity to particle concentration, while electromagnetic signals have strong anti-interference capabilities in identifying particle migration direction. Through fusion analysis, the information carried by the two signals is complementary and integrated to accurately obtain the current concentration of reservoir particles (reflecting the quantity of particles) and their current migration status (including migration direction, migration rate, and whether they are stationary).

[0018] This embodiment of the intelligent tracer and dual-mode sensing CO2 foam injection reservoir particulate monitoring method involves mixing intelligent tracer particles with both fluorescent signal emission and magnetic signal response characteristics with a CO2 foam oil displacement agent in a preset ratio, and then injecting the mixture into the reservoir through an injection well. A dual-mode sensing monitoring device deployed at the reservoir monitoring location collects electromagnetic and fluorescent signals from the reservoir fluid in real time. After preprocessing the signals, including filtering, amplification, and baseline correction, a data fusion algorithm is used to perform collaborative analysis of the preprocessed dual signals. This achieves accurate and real-time monitoring of the current concentration and migration status of reservoir particles under complex environments with high temperature, high pressure, and strong interference, thereby improving extraction efficiency and stability.

[0019] In some embodiments, the process of fusing and analyzing the preprocessed fluorescence signal and the preprocessed electromagnetic signal may include: The fusion strategy is determined based on the salinity of the reservoir fluid and the CO2 foam content. According to the fusion strategy, the preprocessed fluorescence signal and the preprocessed electromagnetic signal are fused and analyzed to obtain the current monitoring data of reservoir particles.

[0020] In a specific implementation, the salinity of reservoir fluids can be obtained indirectly using the electromagnetic sensing module in a dual-mode sensing monitoring device. The propagation characteristics of electromagnetic signals in reservoir fluids are closely related to the fluid salinity; the higher the salinity, the stronger the conductivity of the fluid, and the greater the attenuation of the electromagnetic signal. By pre-establishing the correspondence between the electromagnetic signal attenuation rate and salinity (based on laboratory simulations of electromagnetic signal propagation experiments with fluids of different salinities), the attenuation rate of the pre-processed electromagnetic signal is calculated in real time during monitoring, and the salinity of the reservoir fluid can be determined by comparing it with the corresponding relationship.

[0021] The CO2 foam content can be obtained as follows: It can be assisted by a fluorescence sensing module in a dual-mode sensing monitoring device. CO2 foam scatters fluorescence signals; the higher the foam content, the more pronounced the scattering of the fluorescence signal, and the lower the effective signal transmission intensity. Simultaneously, by combining injection parameters from reservoir development (such as CO2 injection volume and foaming agent concentration), a correlation model between the fluorescence signal scattering coefficient and foam content can be established. By calculating the scattering coefficient of the pre-processed fluorescence signal in real time and calibrating with the injection parameters, the CO2 foam content within the reservoir can be determined.

[0022] After obtaining the salinity and CO2 foam content of the reservoir fluid, these values ​​are compared with preset salinity thresholds and preset foam content thresholds, respectively. The fusion strategy is then determined based on the comparison results. When the mineralization is higher than the preset mineralization threshold and the CO2 foam content is not higher than the preset foam content threshold, the electromagnetic signal is easily distorted by the mineralization, while the fluorescence signal is less affected by interference. At this time, the fusion strategy is based on the fluorescence signal and supplemented by the electromagnetic signal. When the mineralization is not higher than the preset mineralization threshold and the CO2 foam content is higher than the preset foam content threshold, the fluorescence signal is easily distorted by foam scattering, while the electromagnetic signal is less affected by interference. At this time, the fusion strategy is based on the electromagnetic signal and supplemented by the fluorescence signal. When both mineralization and CO2 foam content are not higher than the corresponding preset threshold, the two signals are less affected by interference. At this time, the fusion strategy adopts a weighted average fusion of the two signals. When both mineralization and CO2 foam content are higher than the corresponding preset thresholds, both signals are subject to certain interference. At this time, the fusion strategy adopts adaptive collaborative enhancement fusion.

[0023] After determining the fusion strategy, the preprocessed fluorescence signal and the preprocessed electromagnetic signal are fused and analyzed according to the corresponding strategy: the corresponding signal is used as the main signal, and the effective information related to particle concentration and migration state in the main signal is extracted; the other signal is used as the auxiliary signal to correct the deviation of the main signal or supplement key information. Finally, accurate current monitoring data of reservoir particles are obtained through integration.

[0024] Specifically, when the salinity of the reservoir fluid is determined to be higher than the preset salinity threshold, and the CO2 foam content in the reservoir is not higher than the preset foam content threshold, it is determined that the electromagnetic signal is greatly affected by salinity interference, while the fluorescence signal is less affected. Therefore, the fusion strategy takes the fluorescence signal as the main component and the electromagnetic signal as the auxiliary component. First, the first concentration to be corrected and the current migration state of the reservoir particles are determined based on the preprocessed fluorescence signal. Second, the first concentration to be corrected is corrected based on the trend change of the preprocessed electromagnetic signal, and the final concentration is obtained as the current concentration of the reservoir particles.

[0025] In some embodiments, the process of determining the first concentration to be corrected and the current transport state based on the preprocessed fluorescence signal is as follows: (1) Extracting signal parameters: Extracting fluorescence signal intensity and fluorescence signal lifetime from preprocessed fluorescence signals. Fluorescence signal intensity refers to the amplitude of the signal, which directly reflects the quantity correlation information of particles; fluorescence signal lifetime refers to the time from the emission of the fluorescence signal to its decay to a certain proportion (e.g., 1 / e) of the initial intensity. The fluorescent tracer of the intelligent tracer particles has been calibrated at the factory under interference-free conditions. This reference lifetime is stored in the ground data processing center for subsequent comparison.

[0026] (2) Calculate the first concentration to be corrected: First, the second concentration to be corrected is determined based on a pre-defined mapping relationship between fluorescence signal intensity and particle concentration. This mapping relationship can be pre-calibrated through laboratory experiments: under a simulated reservoir environment (temperature and pressure consistent with the actual reservoir, foam content not exceeding a preset threshold), intelligent tracer particle suspensions of different known concentrations are prepared, and the corresponding fluorescence signal intensities are measured. A one-to-one correspondence between intensity and concentration is established, forming a mapping table stored in the data processing center. During monitoring, the second concentration to be corrected can be obtained by querying the mapping table based on the extracted pre-processed fluorescence signal intensity.

[0027] Then, the second concentration to be corrected is adjusted based on the change in fluorescence signal lifetime relative to the baseline lifetime. In the reservoir environment, factors such as temperature and pressure fluctuations can cause the fluorescence signal lifetime to deviate from the baseline lifetime, thus affecting the accuracy of the mapping between signal intensity and concentration. For example, if the actual signal lifetime is shorter than the baseline lifetime, it indicates that environmental interference has led to an underestimation of the signal intensity, and the second concentration to be corrected needs to be appropriately increased; if the actual signal lifetime is longer than the baseline lifetime, it indicates that the signal intensity has been overestimated, and the second concentration to be corrected needs to be appropriately decreased. During correction, a correction coefficient is determined based on the ratio of the difference between the signal lifetime and the baseline lifetime (the larger the ratio, the larger the absolute value of the correction coefficient). This correction coefficient is used to adjust the second concentration to be corrected, yielding the first concentration to be corrected. This improves the accuracy of concentration calculation and compensates for the interference of environmental factors on the fluorescence signal.

[0028] In some embodiments, the first concentration of reservoir particles to be corrected can also be achieved by the following function:

[0029] in, Indicates fluorescence signal intensity, Indicates fluorescence signal efficiency. This indicates the first concentration of reservoir particles to be corrected. Indicates the optical fiber path length. Indicates the fluid absorption coefficient. This indicates the monitoring radius of the fluorescence signal.

[0030] Fluorescence signal efficiency: This represents the fluorescence intensity produced by a unit concentration of microparticles per unit optical path length. It is related to the fluorescent material properties of the smart tracer microparticles, the excitation wavelength, and the detection sensitivity. It can be obtained through laboratory calibration: Under simulated reservoir temperature and pressure conditions, using a tracer sample of known concentration, its fluorescence intensity is measured, and the efficiency is then calculated by inversely using fiber optic path length and attenuation parameters. value.

[0031] The optical fiber path length refers to the effective path length of the fluorescence signal propagating in the optical fiber sensor, which is determined by the sensor probe design. It is usually determined during sensor manufacturing through optical simulation and actual measurement, and set during deployment based on the probe's insertion depth and structural parameters.

[0032] The fluid absorption coefficient reflects the absorption and attenuation characteristics of reservoir fluids to fluorescence signals, and is related to the fluid composition, salinity, temperature, and pressure. It can be obtained by measuring the absorbance of reservoir fluid samples at fluorescence wavelengths in the laboratory, or by inversion based on historical monitoring data and fluid property models.

[0033] The monitoring radius of a fluorescence signal represents the effective detection range of a fluorescence sensor, and is typically determined by the sensor's sensitivity, the fluorescence signal intensity attenuation pattern, and the background noise level. The monitoring radius can be determined by conducting signal attenuation experiments in simulated geological formations, identifying the furthest distance at which the signal intensity drops to the background noise level.

[0034] (3) Determine the current transport status: First, the intensity variation characteristics and spatial distribution characteristics of the fluorescence signal intensity are obtained. The intensity variation characteristics refer to the evolution of the fluorescence signal intensity over time, including an upward trend (particles gather towards the monitoring area), a stable trend (particles pass through the monitoring area at a constant speed), a downward trend (particles move away from the monitoring area), and fluctuation amplitude (whether there is an alternation of gathering and dispersing). The spatial distribution characteristics refer to the differences in signal intensity collected by different fluorescence sensor nodes, including the intensity gradient between adjacent nodes (the magnitude of the gradient reflects the concentration of particle migration) and the uniformity of the intensity distribution (whether there are local peaks, indicating the location of particle aggregation).

[0035] Then, the two features mentioned above are input into a pre-constructed particle migration pattern classification model for classification. This classification model is built based on machine learning algorithms, and the training process is as follows: Samples of known particle migration states under different reservoir conditions (such as uniform migration, accelerated migration, stagnation, local aggregation, and dispersed migration) are collected. The intensity change features and spatial distribution features corresponding to each sample are extracted. The feature data are then correlated with the corresponding migration state labels to train a classification model that can accurately identify migration states. During monitoring, the model receives the two input features and outputs the corresponding classification result, which is the current migration state of the reservoir particles.

[0036] In some embodiments, the process of correcting the first concentration to be corrected based on the trend change of the preprocessed electromagnetic signal may include: The first step is to set a preset time window. The length of the time window is determined based on the data acquisition frequency and the migration rate of reservoir particles to ensure that the complete trend of signal change can be captured within the window (e.g., 10-30 seconds, which can be adjusted according to actual working conditions).

[0037] The second step is to obtain the intensity sequence of the preprocessed electromagnetic signal corresponding to the first concentration to be corrected within the preset time window: Within the preset time window, the preprocessed electromagnetic signal that is completely synchronized with the fluorescence signal period on which the calculation of the first concentration to be corrected depends is extracted to form an electromagnetic signal intensity sequence (i.e., electromagnetic signal intensity data at each acquisition time within the window); at the same time, the intensity sequence of the preprocessed fluorescence signal used to calculate the first concentration to be corrected is extracted synchronously (i.e., fluorescence signal intensity data at each acquisition time within the window) to ensure that the time dimensions of the two sets of sequences are completely consistent.

[0038] The second step is to extract the signal change trend: The intensity sequence of the preprocessed electromagnetic signal is subjected to trend extraction to obtain the first change trend. Trend extraction can be performed using the moving average method. By setting a sliding window (e.g., 3-5 acquisition cycles), the intensity sequence is smoothed to eliminate short-term fluctuations, preserve long-term evolution patterns, and determine whether the electromagnetic signal intensity shows an upward trend (increasing particle concentration), a downward trend (decreasing particle concentration), or a stable trend (stable particle concentration). Simultaneously, the rate of change of the trend (e.g., the change in intensity per unit time) is recorded.

[0039] The same moving average method was used to extract the trend of the intensity sequence of the preprocessed fluorescence signal to obtain the second trend, which also clarified the evolution direction and rate of change of the fluorescence signal intensity.

[0040] The third step is to calculate trend similarity: Trend similarity measures the degree of agreement between the first and second trends. The calculation considers two dimensions: first, the consistency of the direction of change (both rising, both falling, or both remaining stable indicate a consistent direction; otherwise, they are inconsistent); second, the closeness of the rates of change (calculate the ratio of their rates of change; the closer the ratio is to 1, the closer the rates are). Based on these two dimensions, a similarity scoring standard is set (e.g., if the direction is consistent and the rate ratio is between 0.9 and 1.1, the similarity score is 90-100; if the direction is consistent but the rate ratio is between 0.7-0.9 or 1.1-1.3, the similarity score is 70-89; if the direction is inconsistent, the similarity score is 0-69). The final trend similarity is then determined according to the scoring standard.

[0041] The fourth step is to determine the correction factor and correct the concentration: A correlation between similarity and concentration correction factors is pre-defined, based on experimental calibration: higher similarity indicates a better match between the electromagnetic and fluorescence signal patterns, smaller deviations in the first concentration to be corrected calculated from the fluorescence signal, and a correction factor closer to 1; lower similarity indicates potential undetected interference in the fluorescence signal, and a greater deviation of the correction factor from 1 (e.g., similarity 90-100 points, correction factor 1.0; similarity 70-89 points, correction factor 1.05 or 0.95, adjusted according to the direction of rate difference; similarity 0-69 points, correction factor 1.1 or 0.9, adjusted according to directional difference). Based on the calculated trend similarity, the corresponding correction factor is determined by querying the correlation. This correction factor is then multiplied by the first concentration to be corrected to obtain the corrected concentration, which is the current concentration of reservoir particles.

[0042] This embodiment addresses the specific working conditions of low mineralization and high foam content, fully leveraging the advantages of electromagnetic signals in resisting foam interference. It clarifies the method for calculating the transport state and concentration based on electromagnetic signals. The initial transport direction is identified through magnetic field gradient changes, and the reliability of transport state identification is improved by combining spatial continuity characteristics. The mapping relationship between electromagnetic signal intensity and concentration, along with attenuation model correction, solves the problem of concentration calculation deviation caused by signal propagation attenuation, ensuring the accuracy of the current concentration. The overall solution is adaptable to complex working conditions with high foam content, compensates for the defect of fluorescence signals being affected by scattering interference, and lays a reliable foundation for subsequent transport state correction.

[0043] In some embodiments, when it is determined that the salinity of the reservoir fluid is not higher than a preset salinity threshold and the CO2 foam content in the reservoir is higher than a preset foam content threshold, it is determined that the fluorescence signal is greatly affected by CO2 foam scattering interference and the electromagnetic signal is less affected by salinity interference. Therefore, the fusion strategy takes the electromagnetic signal as the main component and the fluorescence signal as the auxiliary component. First, the first transport state to be corrected and the current concentration are determined based on the preprocessed electromagnetic signal. Second, the first transport state to be corrected is corrected based on the distribution characteristics of the preprocessed fluorescence signal to obtain the final current transport state.

[0044] The process of determining the first transport state to be corrected and the current concentration based on the preprocessed electromagnetic signal is as follows: a1. Obtaining Electromagnetic Signal Intensity Distribution: The electromagnetic sensing module of the dual-mode sensing monitoring device adopts a pre-set spatial array layout of "concentric circles + radial direction," with each magnetic sensor node distributed at a fixed interval. The ground data processing center synchronously receives the pre-processed electromagnetic signals collected by all magnetic sensor nodes, organizes them to obtain signal intensity data corresponding to different spatial locations, and forms the intensity distribution of the pre-processed electromagnetic signals on the pre-set spatial array.

[0045] a2. Calculation of Magnetic Field Gradient Change Characteristics and Identification of Initial Migration Direction: The magnetic field gradient change characteristics reflect the degree and trend of difference in electromagnetic signal intensity between adjacent nodes. During calculation, based on the intensity distribution of the preprocessed electromagnetic signal, the signal intensity difference between adjacent magnetic sensor nodes (including adjacent nodes on the same circumference and adjacent nodes in the radial direction) and the ratio of this difference to the node spacing are calculated sequentially to obtain the magnetic field gradient value for each pair of adjacent nodes. The direction of signal intensity change is determined based on the sign of the gradient value (a positive gradient indicates that the signal of the later node is stronger than that of the earlier node, and a negative gradient indicates that the signal of the later node is weaker than that of the earlier node). The consistent change direction exhibited by most adjacent node pairs is taken as the initial migration direction of the particle (for example, if multiple adjacent nodes in the radial direction all exhibit positive gradients, it indicates that the particle is migrating radially outward).

[0046] a3. Extracting Spatial Continuity Features of Particle Trajectory: Spatial continuity features are used to determine whether the particle migration trajectory is continuous. By analyzing the characteristics of magnetic field gradient changes at different times, we observe the duration of the initial migration direction (e.g., maintaining the same direction for multiple consecutive acquisition cycles indicates high continuity), the fluctuation amplitude of the gradient value (small fluctuation indicates high continuity), and the synergy of gradient changes at different nodes in different regions (e.g., matching radial and circumferential gradient changes indicates high continuity). These indicators are combined to form spatial continuity feature parameters.

[0047] a4. Obtaining the first transport state to be corrected: The pre-trained transport state classification model is built based on a machine learning algorithm. The training samples include known transport states of particles under different operating conditions (such as radial uniform transport, circumferential diffusion transport, local stagnation, and reversible transport), as well as the corresponding preliminary transport direction and spatial continuity features. The preliminary transport direction and spatial continuity features extracted under the current operating condition are input into the model. The model compares the sample features and outputs the corresponding classification result, which is the first transport state to be corrected.

[0048] a5. Calculate the current concentration: First, the third concentration to be corrected is determined based on the preset mapping relationship between electromagnetic signal intensity and particle concentration. This mapping relationship can be obtained through laboratory calibration: under a simulated reservoir environment (mineralization not exceeding a preset threshold, temperature and pressure consistent with the actual reservoir), intelligent tracer particle suspensions of different known concentrations are prepared, and the electromagnetic signal intensity corresponding to different concentrations is measured. A one-to-one correspondence table between intensity and concentration is established and stored in the data processing center. During monitoring, the average signal intensity in the spatial intensity distribution matrix is ​​taken, and the third concentration to be corrected is obtained by querying the mapping table.

[0049] Then, correction is performed using an electromagnetic signal attenuation model in the reservoir medium. This attenuation model is constructed based on parameters such as reservoir medium type (e.g., sandstone, mudstone), temperature, and pressure, and is used to describe the intensity attenuation law of electromagnetic signals propagating in the reservoir (e.g., the greater the medium density and the longer the propagation distance, the more obvious the signal attenuation). Based on the current reservoir geological parameters (e.g., medium type, distance between monitoring nodes and possible particle locations) and production parameters (temperature, pressure), the signal attenuation is calculated by substituting into the attenuation model. It is then determined whether the third concentration to be corrected is underestimated or overestimated due to signal attenuation, and the correction coefficient is determined to adjust the third concentration to obtain the current concentration.

[0050] In some embodiments, the current concentration of reservoir particles can also be achieved using the following function:

[0051] in, Indicates the strength of electromagnetic signals. For tracer magnet density, The volume of the tracer magnet. Indicates temperature-corrected magnetic susceptibility. Indicates porosity. Indicates background interference factor. Indicates the formation magnetic attenuation coefficient. Indicates the monitoring radius of the electromagnetic signal. This indicates the particle concentration of reservoir microparticles.

[0052] The magnet density of a tracer represents the mass-to-volume ratio of magnetic materials (such as iron(II,III) oxide, iron-cobalt alloys, etc.) in a smart tracer microparticle. It is determined by the properties of the magnetic materials themselves and is usually calibrated by the tracer manufacturer or measured in the laboratory using a density measuring instrument (such as the specific gravity bottle method).

[0053] The volume of a tracer magnet refers to the volume of the magnetic portion within a single smart tracer particle. It can be calculated by measuring the particle size using an electron microscope (such as SEM or TEM) and combining this with the percentage of magnetic material, or by a nominal value provided by the manufacturer based on the synthesis process.

[0054] Temperature-corrected magnetic susceptibility represents the magnetization ability of a magnetic material at different temperatures and is significantly affected by temperature. It is necessary to simulate reservoir temperature gradients (e.g., 50–300℃) in a laboratory setting, measure the change in tracer magnetic susceptibility as a function of temperature, establish a temperature-magnetic susceptibility correlation table, or fit it to a temperature function for interpolation based on actual temperature during real-time monitoring.

[0055] Porosity refers to the ratio of pore volume to total volume in reservoir rocks, and is a geological parameter. It can be obtained through well logging interpretation (such as sonic logging, neutron logging, and density logging) or core experimental analysis, and is updated in conjunction with dynamic reservoir models during monitoring.

[0056] Background interference factor reflects the proportion of non-target magnetic interference (such as formation ferromagnetic minerals, downhole tools, etc.) contributing to the magnetic signal in the monitoring environment. A background interference model can be established by acquiring background magnetic signals and analyzing their intensity and distribution characteristics before injecting the tracer, and then subtracting or correcting it from the total signal during real-time monitoring.

[0057] The formation magnetic attenuation coefficient characterizes the ability of the formation medium to absorb and attenuate magnetic signals, and is related to the formation lithology, fluid properties, temperature, and pressure. The attenuation coefficient can be obtained by conducting magnetic signal propagation experiments under simulated formation conditions, measuring the attenuation law of signal intensity with distance, and fitting the data; or by combining historical monitoring data for inversion and optimization.

[0058] The monitoring radius of an electromagnetic signal refers to the radius of the effective detection range of a magnetic sensor, which depends on the sensor sensitivity, magnetic signal strength, and formation attenuation characteristics. It is generally determined through sensor calibration experiments (such as moving a magnetic source in a simulated formation and recording the distance at which the signal attenuates to a threshold) and optimized in conjunction with well network layout during deployment.

[0059] In some embodiments, the process of correcting the first transport state to be corrected based on the distribution characteristics of the preprocessed fluorescence signal may include the following operations: b1. Extracting Fluorescence Intensity Gradient Change Features: The fluorescence sensing module also adopts a "concentric circle + radial" spatial array layout, corresponding one-to-one with the layout of the electromagnetic sensing module. First, the pre-processed fluorescence signals collected by all fluorescence sensor nodes are organized to form a spatial distribution feature matrix of fluorescence signal intensity. Then, referring to the calculation method of electromagnetic signal gradient, the signal intensity difference between adjacent fluorescence sensor nodes (including circumferentially adjacent nodes and radially adjacent nodes) and its ratio to the node spacing are calculated to obtain the fluorescence intensity gradient value of each pair of adjacent nodes. All gradient values ​​together constitute the fluorescence intensity gradient change feature.

[0060] b2. Determine the direction of particle migration trend: Analyze the characteristics of fluorescence intensity gradient changes, and screen out adjacent node pairs with larger absolute gradient values ​​(the larger the gradient value, the more significant the signal contribution of the particles in that direction). Statistically analyze the signal change direction (positive or negative gradient) of these high-gradient node pairs. Determine the direction of change with the highest frequency as the particle migration trend direction corresponding to the fluorescence signal (for example, if most high-gradient node pairs in the radial direction are positive gradients, it indicates that the fluorescence signal indicates that the particles are migrating radially outward).

[0061] b3. Consistency Comparison and Direction Correction: The particle migration trend direction determined by the fluorescence signal is compared with the aforementioned preliminary migration direction to determine whether they are consistent (e.g., both are radially outward or both are circumferentially clockwise, indicating consistency; one is radially outward and the other is radially inward, indicating inconsistency). If inconsistent, correction is made based on the magnitude of the gradient value of the fluorescence intensity gradient change characteristics. First, the average gradient value in the migration trend direction and the average gradient value in the preliminary migration direction are calculated; then, weights are assigned according to the gradient value magnitude (the larger the gradient value, the higher the weight, indicating stronger signal reliability in that direction); finally, the preliminary migration direction is adjusted based on the weights. For example, if the average gradient value in the migration trend direction is twice that in the preliminary migration direction, the migration trend direction is taken as the main factor, combined with the local characteristics of the preliminary migration direction (such as gradient changes at some nodes), to obtain a new migration direction and update the preliminary migration direction.

[0062] b4. Determine the current migration state: If the direction is consistent, directly take the first migration state to be corrected as the current migration state; if the direction has been corrected, re-input the corrected migration direction and the previously extracted spatial continuity features into the migration state classification model, and the model outputs a new classification result, which is the current migration state.

[0063] This embodiment fully utilizes the spatial distribution characteristics of fluorescence signals, enabling the capture of particle migration trends even under conditions with high foam content through gradient analysis. The reliability of the migration trend direction is ensured through gradient calculation of adjacent nodes and high-frequency directional statistics. Consistency comparison and weighted correction mechanisms effectively integrate the directional information of electromagnetic and fluorescence signals, avoiding potential misjudgments from single signals and significantly improving the accuracy of migration state identification. The correction process, combined with the migration state classification model, ensures the integrity and rationality of the corrected migration state, adapting to complex conditions with high foam content.

[0064] In some embodiments, when it is determined that the salinity of the reservoir fluid is not higher than a preset salinity threshold and the CO2 foam content in the reservoir is not higher than a preset foam content threshold, it is determined that both signals are less affected by interference. Therefore, a weighted fusion strategy is adopted to give full play to the advantages of the two signals and obtain accurate current monitoring data.

[0065] Specifically, based on the preprocessed fluorescence signal, a first concentration and a first migration state of the reservoir particles are determined; based on the preprocessed electromagnetic signal, a second concentration and a second migration state of the reservoir particles are determined; the first concentration and the second concentration are weighted and fused to obtain the current concentration; and the first migration state and the second migration state are weighted and fused to obtain the current migration state.

[0066] The process of obtaining the concentration and migration state of reservoir particles for each signal can be referred to the aforementioned relevant records, and will not be repeated here.

[0067] In some embodiments, when it is determined that the salinity of the reservoir fluid is higher than a preset salinity threshold and the CO2 foam content in the reservoir is higher than a preset foam content threshold, an alarm can be triggered. Alternatively, the fusion strategy can be determined to be an adaptive collaborative enhancement fusion strategy, which may specifically include the following operations: c1. Time-frequency analysis and effective signal segment identification: Time-frequency analysis is performed on the preprocessed fluorescence signal and the preprocessed electromagnetic signal respectively (analyzing the intensity distribution of the signal at different times and frequencies). The interference thresholds of the two signals are preset (determined according to the noise level and interference characteristics of the signal itself). The time periods in the signal where the interference level is lower than the corresponding interference threshold are identified. The effective information of the signal is high in the time period, which is regarded as the effective signal segment.

[0068] c2. Cross-validation of valid signal segments: Within the identified valid signal segments, extract the changing trends (e.g., rising, falling, stable) of the preprocessed fluorescence signal and the preprocessed electromagnetic signal, and perform cross-validation. If the two signals show the same trend (e.g., both are slowly increasing), the monitoring results for that period are considered reliable. The preliminary values ​​of concentration and transport state are calculated based on the fluorescence signal using the aforementioned method, and then fine-tuned by combining the aforementioned electromagnetic signal analysis results to obtain the current monitoring data.

[0069] c3. Dynamic weighted fusion when trends are inconsistent: If the trends of the two signals within the effective signal segment are inconsistent, a confidence weight model is constructed by combining reservoir geological parameters (such as porosity and permeability) and production parameters (such as injection pressure and injection volume): The degree of interference of the two signals with the current operating conditions is analyzed (such as the degree of interference of mineralization on electromagnetic signals and the degree of interference of foam on fluorescence signals), and the fusion weight is dynamically allocated (signals with lower interference have higher weights, and signals with higher interference have lower weights); the monitoring results (concentration and migration state) of the two signals are weighted and fused according to the weight to obtain the current monitoring data.

[0070] c4. Adaptive correction when there is no effective signal segment: If no effective signal segment is identified in several consecutive monitoring periods (e.g., 3 periods), or if cross-validation fails, the adaptive correction mechanism is activated: The monitoring data of the previous reliable period (i.e., the period in which the last cross-validation passed) is retrieved, and the changing trend of the current signal and the historical signal is analyzed (e.g., the attenuation or enhancement trend of the current signal relative to the historical signal). Extrapolation compensation is performed based on this trend (e.g., if the historical concentration is X and the current signal trend is a slow increase of 10%, then the extrapolated current concentration is 1.1X). ​​The compensation magnitude is adjusted in combination with the stability of the reservoir conditions, and the current monitoring data is finally output.

[0071] In some embodiments, based on the current monitoring data of reservoir particles, steps for future trend prediction, early warning, and adjustment of mining parameters can be added to form a closed-loop system of "monitoring-prediction-control".

[0072] Specifically, a reservoir particle migration model can be pre-trained based on historical data. Current monitoring data of reservoir particles (current concentration, current migration status), real-time collected reservoir geological parameters (such as dynamic porosity and permeability obtained through well logging equipment), and real-time production parameters (such as dynamic injection pressure and injection volume of injection wells) can be input into the reservoir particle migration model for prediction. The output will be future monitoring data for a preset time period (such as the next 1 hour or 3 hours), including the particle concentration at different future times (future concentration) and the future migration status (such as whether the migration direction will turn towards the wellbore or whether it will accumulate in a certain area).

[0073] In a specific implementation process, a preset early warning threshold can be determined based on reservoir permeability protection requirements. For example, by simulating the impact of different particulate concentrations on reservoir permeability through laboratory experiments, the maximum particulate concentration that will not cause formation damage (permeability reduction not exceeding a preset proportion, such as 5%) can be determined and set as the early warning threshold. The predicted future concentration is compared with the early warning threshold. If the future concentration exceeds the early warning threshold (or it is predicted that it will exceed the early warning threshold at a certain moment), the ground data processing center immediately issues an early warning signal (including audible and visual warnings and data warnings, which are simultaneously sent to the field control terminal and the mobile terminal of the technicians).

[0074] In a specific implementation process, technicians can adjust the mining parameters based on the current monitoring data: if the concentration of particulate matter in a certain area is currently high and the concentration will exceed the standard in the future, the CO2 foam injection pressure in that area will be appropriately reduced to reduce the scouring of reservoir particulate matter by the fluid; if particulate matter is currently detected migrating towards the wellbore and may accumulate in large quantities in the future, a particulate matter stabilizer will be added to the injection agent to inhibit the continued migration of particulate matter; after adjustment, monitoring will continue until the predicted concentration value is lower than the warning threshold.

[0075] Based on the same general inventive concept, this invention also protects an intelligent tracing and dual-mode sensing CO2 foam huff and puff reservoir particulate monitoring system. The intelligent tracing and dual-mode sensing CO2 foam huff and puff reservoir particulate monitoring system provided by this invention will be described below. The intelligent tracing and dual-mode sensing CO2 foam huff and puff reservoir particulate monitoring system described below and the intelligent tracing and dual-mode sensing CO2 foam huff and puff reservoir particulate monitoring method described above can be referred to in correspondence with each other.

[0076] Figure 2 This is a schematic diagram of the intelligent tracing and dual-mode sensing CO2 foam huff and puff reservoir particulate monitoring system provided in an embodiment of the present invention, as shown below. Figure 2 As shown, the intelligent tracing and dual-mode sensing CO2 foam huff and puff reservoir particulate monitoring system of this embodiment includes an injection module 21, an acquisition module 22, a preprocessing module 23, and a fusion module 24.

[0077] The injection module 21 is used to mix intelligent tracer particles with CO2 foam oil displacement agent in a preset ratio and then inject the mixture into the reservoir through an injection well; wherein the intelligent tracer particles include tracers that can emit fluorescent and magnetic signals. The acquisition module 22 is used to monitor the intelligent tracer particles through a dual-mode sensing monitoring device deployed at the reservoir monitoring location, and acquire electromagnetic signals and fluorescence signals from the reservoir fluid. Preprocessing module 23 is used to preprocess the fluorescence signal and the electromagnetic signal to obtain a preprocessed fluorescence signal and a preprocessed electromagnetic signal; The fusion module 24 is used to perform fusion analysis on the preprocessed fluorescence signal and the preprocessed electromagnetic signal using a data fusion algorithm to obtain the current monitoring data of the reservoir particles; the current monitoring data of the reservoir particles includes the current concentration of the reservoir particles and the current migration state of the reservoir particles.

[0078] It should be noted that all relevant information that may be involved in the various embodiments of the present invention is processed in strict accordance with the requirements of laws and regulations, following the principles of legality, legitimacy, and necessity, based on the reasonable purpose of the business scenario, and is information that users actively provide or generate during the use of the product / service, as well as information obtained with user authorization.

[0079] The information processed by this invention may vary depending on the specific product / service scenario and should be based on the specific scenario in which the user uses the product / service. This may involve user account information, device information, or other related information. This invention will treat the relevant information and its processing with the utmost diligence.

[0080] This invention places great emphasis on the security of relevant information and has adopted reasonable and feasible security protection measures that comply with industry standards to protect user information and prevent unauthorized access, public disclosure, use, modification, damage or loss of relevant information.

[0081] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0082] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0083] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for monitoring particulate matter in CO2 foam huff and puff reservoirs using intelligent tracing and dual-mode sensing, characterized in that, include: The intelligent tracer particles are mixed with CO2 foam oil displacement agent in a preset ratio and then injected into the reservoir through an injection well; wherein, the intelligent tracer particles include tracers that can emit fluorescent and magnetic signals; The intelligent tracer particles are monitored by a dual-mode sensing monitoring device deployed at the reservoir monitoring location to obtain electromagnetic and fluorescence signals from the reservoir fluid. The fluorescence signal and the electromagnetic signal are preprocessed to obtain a preprocessed fluorescence signal and a preprocessed electromagnetic signal. The preprocessed fluorescence signal and the preprocessed electromagnetic signal are fused and analyzed using a data fusion algorithm to obtain the current monitoring data of reservoir particles; the current monitoring data of reservoir particles includes the current concentration of reservoir particles and the current migration state of reservoir particles.

2. The method for monitoring particulate matter in CO2 foam huff and puff reservoirs using intelligent tracing and dual-mode sensing according to claim 1, characterized in that, The preprocessed fluorescence signal and the preprocessed electromagnetic signal are fused and analyzed using a data fusion algorithm to obtain the current monitoring data of reservoir particles, including: The fusion strategy is determined based on the salinity of the reservoir fluid and the CO2 foam content. According to the fusion strategy, the preprocessed fluorescence signal and the preprocessed electromagnetic signal are fused and analyzed to obtain the current monitoring data of reservoir particles.

3. The method for monitoring particulate matter in CO2 foam huff and puff reservoirs using intelligent tracing and dual-mode sensing according to claim 2, characterized in that, Based on the salinity and CO2 foam content of the reservoir fluid, a fusion strategy is determined, including: When the salinity of the reservoir fluid is determined to be higher than a preset salinity threshold, and the CO2 foam content in the reservoir is not higher than a preset foam content threshold, it is determined that the electromagnetic signal is interfered with, and the fusion strategy is determined to be: Based on the preprocessed fluorescence signal, the first concentration to be corrected and the current migration state of the reservoir particles are determined; The first concentration to be corrected is corrected based on the trend change of the preprocessed electromagnetic signal to obtain the current concentration.

4. The method for monitoring particulate matter in CO2 foam huff and puff reservoirs using intelligent tracing and dual-mode sensing according to claim 3, characterized in that, Based on the preprocessed fluorescence signal, determining the first uncorrected concentration of the reservoir particles and the current migration state of the reservoir particles includes: Extract the fluorescence signal intensity and fluorescence signal lifetime of the preprocessed fluorescence signal; Based on the preset mapping relationship between fluorescence signal intensity and particle concentration, the second concentration to be corrected of the reservoir particles is determined, and the second concentration to be corrected is corrected by combining the change of fluorescence signal lifetime relative to the reference lifetime under interference-free environment, so as to obtain the first concentration to be corrected. The intensity variation characteristics and spatial distribution characteristics of the fluorescence signal intensity are obtained; The intensity variation characteristics and spatial distribution characteristics of the fluorescence signal intensity are input into a pre-constructed particle transport pattern classification model for classification to obtain the current transport state.

5. The method for monitoring particulate matter in CO2 foam huff and puff reservoirs using intelligent tracing and dual-mode sensing according to claim 3, characterized in that, Correcting the first concentration to be corrected based on the trend change of the preprocessed electromagnetic signal to obtain the current concentration includes: Within a preset time window, the intensity sequence of the preprocessed electromagnetic signal corresponding to the first concentration to be corrected is acquired; Trend extraction is performed on the intensity sequence to obtain the first trend of electromagnetic signal intensity evolution over time; Simultaneously acquire the intensity sequence of the preprocessed fluorescence signal used to calculate the first concentration to be corrected within the same time window, and perform trend extraction to obtain the second trend of fluorescence signal intensity evolution over time; Calculate the trend similarity between the first trend and the second trend; Based on the preset correlation between similarity and concentration correction factor, the correction factor of the first concentration to be corrected is determined; The first concentration to be corrected is corrected using the correction factor, and the corrected concentration is taken as the current concentration.

6. The method for monitoring particulate matter in CO2 foam huff and puff reservoirs using intelligent tracing and dual-mode sensing according to claim 2, characterized in that, The fusion strategy is determined based on the salinity and CO2 foam content of the reservoir fluid, and also includes: When it is determined that the salinity of the reservoir fluid is not higher than a preset salinity threshold, and the CO2 foam content in the reservoir is higher than a preset foam content threshold, it is determined that the fluorescence signal is interfered with, and the fusion strategy is determined to be: Based on the preprocessed electromagnetic signal, the first migration state of reservoir particles to be corrected and the current concentration are determined; The first transport state to be corrected is corrected based on the distribution characteristics of the preprocessed fluorescence signal to obtain the current transport state.

7. The method for monitoring particulate matter in CO2 foam huff and puff reservoirs using intelligent tracing and dual-mode sensing according to claim 6, characterized in that, Based on the preprocessed electromagnetic signal, the first uncorrected transport state of reservoir particles and the current concentration are determined, including: Obtain the intensity distribution of the preprocessed electromagnetic signal on a preset spatial array; Based on the intensity distribution, calculate the magnetic field gradient variation characteristics between adjacent magnetic sensor nodes; Based on the magnetic field gradient variation characteristics, the initial migration direction of reservoir particles is identified, and the spatial continuity characteristics of the migration trajectory are extracted. The initial transport direction and the spatial continuity feature are input into a pre-trained transport state classification model to obtain the first transport state to be corrected. Based on the preset mapping relationship between electromagnetic signal intensity and particle concentration, the third concentration to be corrected of the reservoir particles is determined, and the third concentration to be corrected is corrected by combining the attenuation model of electromagnetic signal in the reservoir medium to obtain the current concentration.

8. The method for monitoring particulate matter in CO2 foam huff and puff reservoirs using intelligent tracing and dual-mode sensing according to claim 7, characterized in that, Correcting the first transport state to be corrected based on the distribution characteristics of the preprocessed fluorescence signal to obtain the current transport state includes: Extract the fluorescence intensity gradient variation characteristics between adjacent fluorescence sensor nodes from the spatial distribution characteristics of fluorescence signal intensity; Based on the fluorescence intensity gradient change characteristics, the direction of particle migration trend corresponding to the fluorescence signal is determined; The consistency of the particle migration trend direction with the initial migration direction is compared: If the directions are consistent, the first movement state to be corrected is taken as the current movement state; If the directions are inconsistent, the initial migration direction is corrected based on the fluorescence intensity gradient change characteristics to obtain the corrected migration direction, and the initial migration direction is updated.

9. The method for monitoring particulate matter in CO2 foam huff and puff reservoirs using intelligent tracing and dual-mode sensing according to claim 2, characterized in that, The fusion strategy is determined based on the salinity and CO2 foam content of the reservoir fluid, and also includes: When it is determined that the salinity of the reservoir fluid is not higher than a preset salinity threshold, and the CO2 foam content in the reservoir is not higher than a preset foam content threshold, the fusion strategy is determined as follows: Based on the preprocessed fluorescence signal, the first concentration of the reservoir particles and the first migration state of the reservoir particles are determined; Based on the preprocessed electromagnetic signal, the second concentration of the reservoir particles and the second migration state of the reservoir particles are determined; The first concentration and the second concentration are weighted and fused to obtain the current concentration, and the first transport state and the second transport state are weighted and fused to obtain the current transport state.

10. The method for monitoring particulate matter in CO2 foam huff and puff reservoirs using intelligent tracing and dual-mode sensing according to claim 1, characterized in that, Also includes: The current monitoring data, reservoir geological parameters, and production parameters are input into a pre-constructed reservoir particle migration model for prediction, thereby obtaining future monitoring data of the reservoir particles; wherein, the future monitoring data includes the future concentration and future migration state of the reservoir particles; If the future concentration exceeds the preset warning threshold, a warning signal will be issued, and the extraction parameters of heavy oil CO2 foam huff and puff will be adjusted according to the current monitoring data.

Images